[unreadable] The goal of the proposed research is to develop, test, and validate a preclinical non-invasive optical instrument, based on spectroscopic Oblique Incidence Reflectometry (OIR), for the rapid diagnosis of pre-cancerous and invasive cancerous skin lesions. The underlying hypothesis, which is well supported by our strong preliminary clinical studies, is that quantitative, spatially resolved, optical-spectral (spatio-spectral) images of skin lesions can provide the dermatologist with clinically valuable diagnostic information to supplement his/her clinical judgment about the need for further diagnostic procedures and/or treatments. OIR intentionally breaks the symmetry of diffuse reflectance that exists in its normal-incidence counterpart and, as a result, provides more robust information about the interrogated tissue volume [Wang et al, US Patent, 5,630,423 (1997)]. A preliminary clinical study of 102 skin lesions using a handheld fiber-optic probe showed that the OIR spectroscopic system differentiated cancerous/pre-cancerous skin lesions from benign lesions with better than 95% accuracy. The applicants have identified the key diagnostic physiological parameters to be (1) the oxygen saturation of hemoglobin, (2) the concentration of total hemoglobin, and (3) the cell-nuclear size. The engineering goal of this proposal is to improve the prototype OIR system to measure skin lesions in real-time and process the data optimally for accurate lesion classification. The mechanistic goal is to identify and quantify the tissue and cellular factors that are responsible for the diagnostic optical features. The clinical goal is to provide accurate, objective, and real-time diagnoses for benign, pre-cancerous, and malignant lesions that are difficult to diagnose clinically. The ultimate societal impact is to improve patient care, save lives, and reduce health care costs. The Specific Aims are listed as follows. 1. To improve instrumentation. 2. To establish a statistically significant database. 3. To develop and validate a diagnostic algorithm. 4. To identify the pathophysiologic parameters responsible for the diagnostic optical features. [unreadable] [unreadable]